Sean Quah created KAFKA-19760:
---------------------------------
Summary: RecordTooLargeExceptions in group coordinator when
offsets.topic.compression.codec is used
Key: KAFKA-19760
URL: https://issues.apache.org/jira/browse/KAFKA-19760
Project: Kafka
Issue Type: Bug
Components: group-coordinator
Reporter: Sean Quah
The coordinator runtime can be overly optimistic about compression ratios when
filling up batches and can often overflow `max.message.bytes` of the topic.
When this happens, all the group coordinator requests that wrote to the batch
fail with an UNKNOWN_SERVER_ERROR. This error is non-retriable and disruptive
to client applications.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
